Do abstracts in otolaryngology randomized controlled trials (RCTs) contain spin, the misrepresentation and distortion of research findings?
Bottom Line: Spin was common in this sample of otolaryngology RCTs.
Background: RCT abstracts play an important role in clinical decision making and should portray study findings as accurately and succinctly as possible. Although there are strict regulations regarding research conduct, authors can choose how their findings are reported, especially in abstracts, which can intentionally or unintentionally bias reporting, often toward statistically significant findings and conclusions.
Study design: Cross-sectional analysis of 47 RCTs with statistically nonsignificant results for primary outcomes.
Setting: PubMed database.
Synopsis: Most studies evaluated a pharmacological or surgical intervention against active treatments; about a third compared against a placebo or sham procedure. Eleven trials were funded by industry, while nearly half did not mention a funding source. (Spin was detected in 15 of these studies’ abstracts.) The primary outcome was clearly stated in either the results or methods section of 28 abstracts, while 20 had no reported numerical value of significance for the primary outcome (although further analysis revealed that 16 described a nonsignificant finding regarding the primary outcome). Spin was identified in 33 abstracts, but was not present in study titles; of the 33, 15 reported a trial registration number. Spin was found in the results sections of 25 abstracts; the most common strategy was focusing on a positive secondary endpoint. Spin was found in the conclusion sections of 27 abstracts; the most common strategy was claiming equivalence based on a negative endpoint. Spin due to reporting bias was not detected in any studies.
Citation: Cooper CM, Gray HM, Ross AE, et al. Evaluation of spin in the abstracts of otolaryngology randomized controlled trials. Laryngoscope. 2019;129:2036–2040.